Tumbo et al. Virol J (2021) 18:28 https://doi.org/10.1186/s12985-021-01500-8

RESEARCH Open Access Role of human Pegivirus infections in whole Plasmodium falciparum sporozoite vaccination and controlled human malaria infection in African volunteers Anneth‑Mwasi Tumbo1,2,3, Tobias Schindler2,3, Jean‑Pierre Dangy2,3, Nina Orlova‑Fink2,3, Jose Raso Bieri4, Maximillian Mpina1,2,3,4, Florence A. Milando1, Omar Juma1, Ali Hamad1,4, Elizabeth Nyakarungu1,4, Mwajuma Chemba1,4, Ali Mtoro1,4, Kamaka Ramadhan1,4, Ally Olotu1,4, Damas Makweba9,10,11, Stephen Mgaya10,11, Kenneth Stuart5, Matthieu Perreau6, Jack T. Stapleton7, Said Jongo1,4, Stephen L. Hofman8, Marcel Tanner2,3, Salim Abdulla1,4 and Claudia Daubenberger2,3*

Abstract Background: Diverse vaccination outcomes and protection levels among diferent populations pose a serious challenge to the development of an efective malaria vaccine. Co-infections are among many factors associated with immune dysfunction and sub-optimal vaccination outcomes. Chronic, asymptomatic viral infections can contribute to the modulation of vaccine efcacy through various mechanisms. Human Pegivirus-1 (HPgV-1) persists in immune cells thereby potentially modulating immune responses. We investigated whether Pegivirus infection infuences vaccine-induced responses and protection in African volunteers undergoing whole P. falciparum sporozoites-based malaria vaccination and controlled human malaria infections (CHMI). Methods: HPgV-1 prevalence was quantifed by RT-qPCR in plasma samples of 96 individuals before, post vaccina‑ tion with PfSPZ Vaccine and after CHMI in cohorts from Tanzania and Equatorial Guinea. The impact of HPgV-1 infec‑ tion was evaluated on (1) systemic cytokine and chemokine levels measured by Luminex, (2) PfCSP-specifc antibody titers quantifed by ELISA, (3) asexual blood-stage parasitemia pre-patent periods and parasite multiplication rates, (4) HPgV-1 RNA levels upon asexual blood-stage parasitemia induced by CHMI. Results: The prevalence of HPgV-1 was 29.2% (28/96) and sequence analysis of the 5′ UTR and E2 regions revealed the predominance of genotypes 1, 2 and 5. HPgV-1 infection was associated with elevated systemic levels of IL-2 and IL-17A. Comparable vaccine-induced anti-PfCSP antibody titers, asexual blood-stage multiplication rates and pre- patent periods were observed in HPgV-1 positive and negative individuals. However, a tendency for higher protection levels was detected in the HPgV-1 positive group (62.5%) compared to the negative one (51.6%) following CHMI. HPgV-1 viremia levels were not signifcantly altered after CHMI.

*Correspondence: [email protected] 2 Department of Medical Parasitology and Infection Biology, Clinical Immunology Unit, Swiss Tropical and Public Health Institute, Socinstr. 57, 4002 Basel, Switzerland Full list of author information is available at the end of the article

© The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://crea‑ tivecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdo‑ main/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Tumbo et al. Virol J (2021) 18:28 Page 2 of 18

Conclusions: HPgV-1 infection did not alter PfSPZ Vaccine elicited levels of PfCSP-specifc antibody responses and parasite multiplication rates. Ongoing HPgV-1 infection appears to improve to some degree protection against CHMI in PfSPZ-vaccinated individuals. This is likely through modulation of immune system activation and systemic cytokines as higher levels of IL-2 and IL17A were observed in HPgV-1 infected individuals. CHMI is safe and well tolerated in HPgV-1 infected individuals. Identifcation of cell types and mechanisms of both silent and productive infection in individuals will help to unravel the biology of this widely present but largely under-researched virus. Keywords: Malaria, Human pegivirus, Controlled human malaria infection, Immune activation, Antibody response, PfSPZ vaccine

Background Human Pegivirus-1 (HPgV-1), a predominantly asymp- Vaccination is an invaluable tool in public health that tomatic virus causing a chronic infection, is common has contributed to control of many, and in some cases, in Africa where an estimated 18–28% of its roughly 750 to the elimination of infectious disease like smallpox million global infections occur [18]. Te virus establishes [1]. Malaria, a disease caused by Plasmodium species its persistence potentially by replicating in immune cells remains a major public health burden particularly in including T cells, B cells, monocytes, and natural killer the tropics and sub-tropical regions where it accounted (NK) cells [19, 20]. Interestingly, seminal feld studies for approximately 405,000 deaths in 2018 [2]. A major have linked HPgV-1 co-infection status to signifcant goal in malaria research is to develop an efcacious survival advantages in HIV-1 and Ebola infected humans vaccine that complements currently used control [21–24]. Tese favourable outcomes are thought to be tools based on vector control and treatment of clinical based on immune-modulatory properties of HPgV-1 malaria infections [3]. However, these vaccine develop- such as reduced activation of T cells, B cells and NK ment eforts are challenged by an incomplete under- cells [20, 25] and the altered regulation of cytokine and standing of the immune mediators leading to highly chemokine expression [26–28]. Diferent HPgV-1 geno- protective, long-lasting vaccine induced immunity in types might infuence the extent of immune modulation the feld [4]. A number of studies testing cryopreserved, resulting in varied disease outcomes [21–23]. purifed, metabolically active and radiation-attenuated Given the high prevalence of HPgV-1 infection in P. whole sporozoites of P. falciparum as vaccine approach falciparum endemic countries, we expected an overlap- (PfSPZ Vaccine) have been published recently [5–9]. ping geographical distribution and aimed to investigate Strikingly, the comparison of PfSPZ vaccine-induced within-host interactions between the two infections. antibody titers specifc for the P. falciparum circum- We were therefore interested to study whether HPgV-1 sporozoite protein (PfCSP) showed signifcantly lower infection status might infuence PfSPZ vaccine-induced titers in malaria pre-exposed than malaria-naive indi- immune responses. We characterized prevalence and viduals immunized with the PfSPZ Vaccine using com- genotype distribution of HPgV-1 in three cohorts of adult parable regimen [6–9]. Tese diferences in PfSPZ volunteers participating in PfSPZ Vaccine studies [9, 29, vaccine-induced immunity was also observed between 30]. We explored the infuence of HPgV-1 infection status vaccinees residing in malaria endemic countries includ- on cytokine and chemokine levels in serum and corre- ing Tanzania, Mali and Equatorial Guinea [10–12]. lated HPgV-1 infection on vaccine-induced anti-PfCSP- Recently our group demonstrated that age, location antibody titers and protection against homologous and iron status infuence the immune system develop- CHMI. We also aimed to characterize for the frst time ment of children as well as vaccine-induced responses the potential impact of a CHMI study on HPgV-1 viremia to the most advanced malaria vaccine candidate, the in these volunteers. RTS,S [13]. Additionally, communicable and non-com- municable diseases have been implicated in suboptimal Methods vaccine-induced responses [14] Chronic, asympto- Study population matic viral infections at time of immunization might We used samples from volunteers enrolled in four studies contribute to reduced magnitude and longevity of vac- conducted in Bagamoyo, Tanzania (acronyms BSPZV1, cine-induced immune responses [15–17]. To date, the BSPZV2, and BSPZV3a) and on Bioko Island in Equa- number of human viruses investigated in this context torial Guinea (acronym EGSPZV2) registered at Clini- is limited and their mechanisms in modulation of vac- caltrials.gov with registration numbers NCT03420053, cine-induced responses remain unclear. NCT02132299, NCT02613520, and NCT02859350, respectively. Detailed trial designs and study procedures Tumbo et al. Virol J (2021) 18:28 Page 3 of 18

such as pre-defned inclusion and exclusion criteria have from commonly used viral metagenomic analytical tools been described previously [9, 10, 29, 30]. Briefy, these tri- [32–36]. Briefy, our analyses were carried out in three als evaluated the safety, immunogenicity and efcacy of consecutive steps: viral identifcation, in silico valida- live, cryopreserved, purifed, whole P. falciparum sporo- tion and RT-PCR confrmation. In the viral identifcation zoites in malaria pre-exposed volunteers. Te analyses in step, we analysed approximately 3 million non-human, these studies were performed on samples collected from unmapped paired end reads from each volunteer. Te adult volunteers in which vaccine efcacy was evalu- initial reads were frst searched for “suspected” viral hits ated by homologous CHMI based on direct intravenous by running bowtie2 against the NCBI database contain- inoculation of 3200 fully infectious, aseptic purifed ing more than 7424 viral genomes. Tereafter, low quality cryopreserved P. falciparum sporozoites. Te current and complexity reads as well as reads mapping to human study was performed in two parts (i) a pilot virome study genome, transcriptome and repeat regions were removed that included samples from a subset of volunteers from from the resulting “suspected” viral reads using bowtie2, BSPZV1 (NCT03420053) (Additional fle 1: Fig. 1); (ii) knead data and tandem repeat fnder algorithms, respec- the main study which utilized samples from volunteers tively. Te “clean” viral reads were then comprehensively participating in the BSPZV2 (NCT02132299), BSPZV3a searched for viral hits using virome scan [32] and Tax- (NCT02613520) and EGSPZV2 (NCT02859350) trials. onomer [33] and for viral proteins using adapted Dia- Samples were selected based on availability and scien- mond tool containing a custom made database with more tifc aims. Tested sample types and sizes are described than 100,000 viral proteins [35]. Te initial unmapped in further detail in each section and in Additional fle 1: reads were also analysed by Fast virome explorer with- Fig. 1A. out fltering for host reads to allow the identifcation of endogenous retroviral elements and other viruses that Identifcation of human Pegivirus RNA in RNA‑seq data may have been missed previously [34]. Only viral hits from whole blood known to be associated with a human host were selected, Whole venous blood samples were used from a subset of and viral contaminants such as lymphotropic murine participants (n = 28) (Additional fle 1 Fig. 1A) participat- virus and synthetic constructs coding for either HIV-1 ing in the BSPZV1 vaccine trial (NCT03420053) based or hepatitis B were removed based on documented lit- on their availability. All volunteers were healthy males, erature [37]. In a following in-silico confrmatory step, aged 18 to 35 years and confrmed as negative for HIV-1, the suspected viral hits were blasted and mapped against Hepatitis B and C at enrolment. Venous blood was col- specifc viral whole genomes using a Geneious bioinfor- lected at diferent time points including: before vaccina- matics tool [38]. As a last step, we performed reverse tion (baseline), 2 and 7 days after frst vaccination, 7 days transcription PCR (RT-PCR) analysis. Due to limited after the second vaccination, before CHMI, 2 and 9 days sample availability we were unable to screen the entire after CHMI. Each of the placebo (n = 6) and the vaccine BSPZV1 cohort for HPgV-1. Hence, the presence of the (n = 22) participants had a total of 3 and 7 blood sampling most prevalent virus (human pegivirus-1, HPgV-1) was time points screened respectively, resulting in 172 sam- confrmed by RT-PCR in plasma samples of volunteers ples in total. All blood samples (n = 172) were stored in found positive by RNA-seq transcriptome analysis only. Paxgene RNA tubes and subjected to RNA-seq analysis performed as published [31]. Briefy, RNA-seq data was RT‑qPCR for detection and quantifcation of HPgV‑1 generated from whole blood RNA (depleted for globin/ and HPgV‑2 rRNA) that was fragmented; the frst strand cDNA syn- Plasma samples collected from male and female indi- thesis was done by random priming and dTTP was used, viduals (n = 96), aged 18–45 years, and participating in whereas 2nd strand cDNA synthesis used dUTP which the BSPZV2, BSPZV3a and the EGSPZV2 studies were eliminates 2nd strands in the downstream PCR amplif- included. Plasma was prepared by density gradient cen- cation that enabled strand specifc RNA-seq sequencing trifugation of whole blood and cryopreserved. At analysis [31]. From the RNA-seq sample set (n = 172), 800 million cryopreserved plasma was thawed and used for detection non-human reads were identifed. Given the naturally of HPgV-1 and HPgV-2 RNA in all study participants. occurring fuctuation of viremia of many viruses, we per- Plasma samples collected at 3 time points for each vol- formed a longitudinal assessment of viral infection sta- unteer were included, namely before vaccination (base- tus, and obtained reads from all available time points for line), before CHMI and 28 days after CHMI. Presence or each of the volunteers. Te analyses was performed in an absence of HPgV-1 and HPgV-2 was determined simul- in-house developed viral metagenomics analysis pipeline taneously using RT-qPCR based on published methods outlined in (Additional fle 1: Fig. 1B–C). Te pipeline is [25]. Briefy, total nucleic acids were extracted from 300 a combination of several published algorithms adapted ul plasma using Zymo quick DNA/RNA viral kit (Zymo Tumbo et al. Virol J (2021) 18:28 Page 4 of 18

Fig. 1 Unbiased search for RNA molecules encoding human viruses in RNA-seq transcriptomics data. a Overall prevalence of 9 human viruses detected in 172 whole blood samples b Number of viral RNA-seq reads detected for each of the identifed viruses. Virus names are plotted on the y-axis and proportion (a), number of reads (B) on the x-axis. c Distribution of the 9 diferent viruses across the 28 individuals included. Virus names are plotted on the y-axis and volunteer IDs on the x-axis. Each bar indicates viral reads for an individual. The log viral RNA-seq reads are plotted, in increasing order ranging from 0 to 3; green indicating low number and red high number of reads

Research, Irvine, USA) and eluted in 50 ul of DNase/ Genotyping of HPgV‑1 RNase free water. 5 ul of the recovered DNA/RNA solu- Te Fire Script cDNA kit was used to synthesize cDNA tion was used as amplifcation template together with in accordance to manufacture instructions (Solis Bio- 2X.Lunar universal one step qPCR master mix (10 ul, dyne, Tartu, Estonia). Briefy, 5 ul of extracted RNA as 1X), Luna warm start reverse transcriptase enzyme mix described above was added into a master mix contain- (1 ul, 1X) (New England Biolabs, MA, USA) and prim- ing forward and reverse primers specifc to 5′ UTR of ers binding to the 5′ untranslated regions of HPgV-1 HPgV-1 (each at 1,1 uM), deoxribonucleotide triphos- and HPgV-2 (each at 2 ul, 0.4 uM) [25, 39]. In addition, phate mix (dNTP) (0.5 ul, 500 uM), reverse transcription human RNaseP primers were added as internal control. bufer with DTT (2 ul, X1), RiboGrip Rnase inhibitor (0.5 Each sample was run in triplicate in a one-step multiplex ul, 1 U/ul), Fire script reverse transcriptase (Solis Bio- RT-qPCR using the CFX96 real time PCR system (Bio- dyne, Tartu, Estonia) (2 ul, 10 U/ul) and RNase free water Rad, Hercules, CA, USA). Te RT-qPCR cycling condi- (9 ul to 20 ul). Amplifcation conditions included 50 min tions were: 55 °C for 10 min, 95 °C for 1 min, 45 cycles at 50 °C and 10 min at 94 °C. 3 ul of cDNA generated by at 95 °C for 15 s and 55 °C for 1 min. Te generated data reverse transcription were used for the frst round of PCR were uploaded to an in-house available analysis platform amplifcation with forward primer 5′-AAA​GGT​GGT​ where quantifcation cycle values (Cq) were calculated GGA​TGG​GTG​ATG-3′ [41] and reverse primer 5′-ATG automatically [40]. HPgV viral quantifcation was done CCACCC​ GCC​ CTC​ ACC​ AGA​ A-3​ ′ combination [41]. as described by Stapleton et al. using in vitro transcribed 1.2 ul of this amplifcation product was then used for (IVT) viral RNA [25]. In each RT-qPCR experiment, we the nested PCR amplifcation using the forward primer included a positive (HPgV-1 IVT-RNA), a negative (from 5′-AAT​CCC​ GGT​CAY​AYT​GGT​AGC​CAC​T-3′ and HPgV-1 negative volunteer) and a non-template control. reverse primer 5′-CCC​CAC​TGGCZTTG​YCA​ACT-3′ Tumbo et al. Virol J (2021) 18:28 Page 5 of 18

combination [41]. Both PCR reactions included primers ProcartaPlex™ Panel 1 (Afymetrix Biosciences, USA) specifc for HPgV-1 (1 ul, 1 uM), frepol master mix (4 ul, and acquired on a validated Luminex XMAP technology X5) (Solis Biodyne, Tartu, Estonia) and RNase free water platform as described [43]. Te investigated cytokines to a fnal volume of 20 ul. Cycling conditions were 5 min and chemokines included BDNF, Eotaxin/CCL11, EGF, at 95 °C, followed by 28 cycles of 95 °C for 30 s, 56 °C for FGF-2, GM-CSF, GRO alpha/CXCL1, HGF, NGF beta, 30 s and 72 oC for 30 s with a fnal extension step at 72 °C LIF, IFN alpha, IFN gamma, IL-1 beta, IL-1 alpha, IL- for 10 min. Te E2 region was amplifed as described by 1RA, IL-2, IL-4, IL-5,IL-6, IL-7, IL-8/CXCL8, IL-9, IL-10, Souza et al. [42]. Te fnal PCR products from 5′ UTR IL-12 p70, IL-13, IL-15, IL-17A, IL-18, IL-21, IL-22, amplifcation (256 base pairs) and E2 amplifcation (347 IL-23, IL-27, IL-31, IP-10/CXCL10, MCP-1/CCL2, MIP-1 base pairs) were sequenced by the Sanger sequencing alpha/CCL3, MIP-1 beta/CCL4, RANTES/CCL5, SDF-1 method (Microsynth, Switzerland). alpha/CXCL12, TNF alpha, TNF beta/LTA, PDGF-BB, PLGF, SCF, VEGF-A and VEGF-D. Only cytokines and HPgV‑1 phylogenetic analysis chemokines with levels above the pre-defned lower Nucleotide sequence analysis and phylogenetic analy- detection limit of the specifc standard curves were sis was performed with the Geneious software version included in the group comparisons. Absolute concentra- 8.1.9. Chromatograms were examined for quality, and tions were normalized to account for the inter-plate vari- only sequences with quality threshold above 86% were ations before analysis in R software version 3.5.1. included in analysis. CLUSTALW algorithm was used to align 5` UTR nucleotide sequences from volunteers to Serological analysis selected reference sequences corresponding to 5` UTR Serum samples for anti-PfCSP antibody evaluation were of HPgV-1 (genotype 1 to 7) available through the NCBI collected before vaccination (baseline) and 14 days post database. Tereafter, phylogenetic trees were constructed last vaccination. Anti-PfCSP total IgG levels were meas- by the neighbour joining method and the Kimura two ured by enzyme linked immunosorbent assay (ELISA) as parameter models. Te references sequences for 5` UTR described previously [9, 10, 29, 30]. of HPgV-1 included AF488786, AF488789, KC618399, KP710602, U36388, JX494177, Y16436, and MF398547 Quantitative detection of Plasmodium falciparum (Genotype 1); AB003289, AF104403, D90600, JX494179, Asexual blood-stage malaria parasitemia during CHMI MG229668, JX494180, U4402, U59518 (Genotype 2; 2a), was assessed using thick blood smear (TBS) microscopy MH000566, U59529, U63715, MH053130 (Genotype 2; and retrospectively analysed using stored whole blood 2b); AB008335, KR108695, JX494176, D87714 (Geno- samples and qPCR as described [9, 10, 29, 30]. Whole type 3); AB0188667, AB021287, HQ3311721 (Genotype blood samples for the assessment of parasitemia were 4); DQ117844, AY949771, AF488796, AF488797 (Geno- taken before CHMI and during the observation period type 5); AB003292, AF177619 (Genotype 6), HQ331235, beginning at day 9 after parasite challenge inocula- HQ 3312233 (Genotype 7). Te hepatitis C nucleotide tion until volunteers either became asexual blood-stage sequence deposited under AJ132997 was used as an out- malaria positive or until day 21. TBS were performed group. For the E2 region the sequences were KP701602.1, twice a day from day 9 to 14 and then once a day for KM670109, U36380, KP710600, KC618399, AB003291 day 15 to 21. TBS were also performed on day 28 before (Genotype 1); AF121950, MK686596, D90600 (Geno- malaria drug treatment. Pre-patent periods were calcu- type 2a) and U63715 (Genotype 2b); D87714 (Genotype lated from the time of PfSPZ challenge to frst positivity 3); AB0188667 (Genotype 4); AY949771, KC618401, detected by qPCR and TBS [9, 10, 29, 30]. Parasite mul- AY951979 (Genotype 5); AB003292 (Genotype 6). A tiplication rate (PMR) was assessed using a linear model Chimpanzee HPgV-1 strain deposited under AF70476 ftted to log10-transformed qPCR data as published [44]. was used as an out-group and U4402 (Genotype 2) was PMR was calculated for all volunteers that developed used for mapping of our sequences to identify regions of asexual blood-stage parasitemia which lasted for at least similarity. two 48-h cycles [44].

Ex vivo cytokine and chemokine measurements Statistical analysis Serum samples available from 44 volunteers collected Figures and statistical analyses were generated in R ver- from EGSPZV2 and BSPZV3a (only HIV-1 negative sion 3.5.1 and GraphPad Software (Prism V5). Wil- volunteers) at baseline were used for the assessment of coxon rank sum test or Mann–Whitney test were used the systemic immune activation status. Cytokine and to compare continuous variables. Chi-square test was chemokine concentrations were measured using the used to compare categorical variables. Absolute values Cytokine/Chemokine/Growth Factor 45-Plex Human for antibody titers and concentrations of cytokines and Tumbo et al. Virol J (2021) 18:28 Page 6 of 18

chemokines were plotted. Data were log transformed of 8 volunteers the in silico identifed presence of HPgV-1 only when investigating the anti-PfCSP antibody titres RNA. Interestingly, these 2 volunteers had the highest and viremia levels. Spearman correlation was used to RNA read counts for HPgV-1 in our bioinformatics anal- investigate the potential efect of HPgV-1 infection status ysis (Fig. 1c). Tese results were important for selecting and viremia with antibody titres and cytokine levels. Data HPgV-1 as our further research focus and the optimi- for cytokines, chemokines and growth factors were not zation of RT-PCR assay used for assessment of HPgV-1 analysed for multiple correction as we considered this infection status in the main study described below. question as exploratory. P value ≤ 0.05 was considered signifcant. Diferences in viral diversity, abundance and Prevalence of HPgV‑1 in East and West African volunteers prevalence were assessed using Linear discriminant anal- After having established that HPgV-1 is highly present in ysis efect size [45] and GraphPad Software (Prism V5), our Tanzanian cohort (BSPZV1), we aimed to explore the respectively. prevalence of HPgV-1 and HPgV-2 in two larger cohorts from Tanzania (BSPZV2 and BSPZV3a) and one from Results Equatorial Guinea (EGSPZV2). Plasma samples col- Unbiased search for RNA molecules encoding human lected from 96 participants, including 12 HIV-1 posi- viruses tive individuals, were analysed for presence of HPgV-1 We aimed to identify viruses present in peripheral blood and HPgV-2 using optimized RT-qPCR. Te overall of our volunteers participating in PfSPZ Vaccine studies. prevalence of HPgV-1 was 29.2% (28/96) (Fig. 2a), while Tese analyses included samples from 28 participants HPgV-2 was not detected. Te proportion of HPgV-1 of the BSPZV1 study collected at multiple time points positive individuals by gender and geographic location including baseline, 2 days after frst vaccination, 7 days were comparable, with slightly more HPgV-1-positive after the frst and second vaccination and before CHMI, individuals in Equatorial Guinea (31.4%) than Tanzania 2 and 9 days after CHMI. Sequences were identifed (26.7%) (Fig. 2b, c). Of the 12 HIV-1 positive individuals from a pool of RNA-seq data reads that did not map to from the BSPZV3a study, two (16.7%, 2/12) were positive the human reference transcriptome. A total of 800 mil- for HPgV-1 (Fig. 2d). lion non-human RNA-seq reads derived from 172 whole blood samples were analysed with our virome discovery HPgV‑1 viral loads and distribution platform based on previously established metagenomics Next, we quantifed the HPgV-1 viral load in plasma sam- pipelines and tools (Additional fle 1: Fig. 1B, C) [32–35]. ples using RT-qPCR. HPgV-1 viral loads were compara- In total, RNA molecules encoding 9 human viruses ble between individuals from the two countries (Fig. 3a). were detectable including the Human simplex virus However, based on viral loads with a predefned thresh- (HSV-1), Cytomegalovirus (CMV), Epstein-Barr virus old of 10­ 6 viral RNA copies/ml of plasma, both cohorts (EBV), Merkel cell polyomavirus (MCV), Human mast could be divided into HPgV-1 low and high viremic indi- adenovirus (HAdV), Astrovirus MBL2, Human beta- viduals (Fig. 3b). High and low HPgV-1 viremia were herpesvirus 7 (HHV-7), Human endogenous retrovirus found in 17 (60%) and 11 (40%) of the 28 HPgV-1 positive K113 (HERV-K113) and HPgV-1 (Fig. 1a). Te number volunteers, respectively (Fig. 3b). Of the 17 high viremic of reads for each of the identifed viruses was quanti- individuals, 8 were from Tanzania and 9 from Equatorial fed and is given in Fig. 1b. After identifying 9 diferent Guinea. Of the 11 low viremic individuals, 4 were Tanza- viruses present in 172 whole blood samples, we further nians and 7 Equatorial Guineans. assessed the distribution of viruses within our cohort. HERV-K113 was detected with high number of reads in Genotyping of HPgV‑1 isolates all 28 individuals, while HSV-1 and CMV were present Seven diferent genotypes of HPgV-1 have been described in seven and six individuals, respectively (Fig. 1c). MBL2, globally with genotype 1 and 5 being highly prevalent in HHV-7 and HAdV were present in low read counts in sub-Saharan Africa [26]. Terefore, we determined the one individual each, and MCV was found in two indi- phylogenetic relatedness of the isolates by amplifying and viduals. Eight individuals carried HPgV RNA with read sequencing the 5′ UTRs. From the 28 positive individu- counts ranging from low to high (Fig. 1c). Tree out of als, 2 were excluded due to poor quality of the sequences 8 HPgV-positive individuals were co-infected with CMV obtained. Genotype 1 was found in 2 volunteers (7.7%) (Fig. 1c). Our analysis showed that a high proportion of and surprisingly, genotype 2, described as dominant in Tanzanian adults (28.6%, 8/28) harboured HPgV -1 infec- Europe and America, was found in 24 of 26 volunteers tion. To reconfrm our fndings, we extracted RNA from (92.3%) (Fig. 4). Most genotype 2 strains clustered closely plasma samples collected from these 8 volunteers and with the related genotype 2a sequences described from amplifed HPgV-1 by RT-PCR. We reconfrmed in 2 out Venezuela (Fig. 4). To further increase the resolution of Tumbo et al. Virol J (2021) 18:28 Page 7 of 18

Fig. 2 Proportion of individuals with (purple) and without (grey) HPgV- 1 infection. a Total cohort of 96 vacinees, b separated by gender, c Country of origin, d HIV-1 infection status. All individuals are between 18 and 35 years of age. Chi square with Yates correction for group comparisons (*P < 0.05)

Fig. 3 Comparisons of HPgV-1 viral loads. No diferences in HPgV-1 viral loads between Equatorial Guinea (green, n 16) and Tanzania (blue, = n 12) volunteers (a). Two distinct groups with low (blue) and high (grey) viremia levels in plasma are found in HPgV-1 infected individuals (b). The = two groups were divided based on a cut of value of 600,000 RNA copies/ml plasma the genetic relatedness of our isolates, we amplifed in circulating in Tanzania and Equatorial Guinea, clustering addition the polymorphic E2 region of HPgV-1. E2 RNA to published genotypes 1, 2 and 5. was successfully amplifed and sequenced in 9 of the 28 positive volunteers (32%). According to the E2-sequences Efect of HPgV‑1 positivity on systemic cytokine of our HPgV-1 isolates, our strains clustered within gen- and chemokine levels otypes 1, 2, and 5 (Fig. 5). Notably, genotypic cluster- To dissect whether ongoing HPgV-1 infection afects ing of isolates 6EG and 14EG difered based on E2 and cytokine and chemokine levels in serum, we measured 45 5′ UTR derived PCR products (Figs. 4, 5). In summary, cytokines, chemokines and growth factors in a subset of these results show that a range of HPgV-1 genotypes are 44 volunteers from BSPZV3a (HIV-1 negative only) and Tumbo et al. Virol J (2021) 18:28 Page 8 of 18

Fig. 4 Phylogenetic inferences of the HPgV-1 isolates based on 5′ UTR. Phylogenetic tree was constructed using Neighbour joining method and Kimura two-parameter model of the 5′ UTR. The 5′ UTR sequences from Tanzania and Equatorial Guinea (n 26) were compared to selected = references spanning genotype 1 to 7 from diferent countries available in the NCBI database. The accession numbers for the reference sequences were: AF488786, AF488789, KC618399, KP710602, U36388, JX494177, Y16436, and MF398547 (Genotype 1, Pink); AB003289, AF104403, D90600, JX494179, MG229668, JX494180, U4402, U59518 (Genotype 2; 2a light brown), MH000566, U59529, U63715, MH053130 (Genotype 2; 2b Brown); AB008335, KR108695, JX494176, D87714 (Genotype 3, Green); AB0188667, AB021287, HQ3311721 (Genotype 4, Maroon); DQ117844, AY949771, AF488796, AF488797 (Genotype 5, Light blue), AB003292, AF177619 (Genotype 6, Bright green); HQ331235, HQ 3312233 (Genotype 7, Golden) and Hepatitis C (AJ132997, Black) was used as outgroup

EGSPZV2. 23 cytokines, chemokines and growth fac- the presence of HPgV-1 infection increases IL-2 and IL- tors were detected above their pre-defned lower limits 17A levels in peripheral blood. of detection (Additional fle 5: Table 1). Although there was a trend of overall higher cytokine levels in HPgV-1 Efect of HPgV‑1 infection status on PfSPZ vaccine‑induced infected individuals (Additional fle 2: Fig. 2), only IL-2 humoral immune response and IL-17A reached signifcance levels (Fig. 6). Tere IL-2 and IL-17A might contribute to diferentiation of was no statistically signifcant diference in cytokine and naïve B cells into plasma cells and support the survival chemokine levels when HPgV-1 high and low viremic of activated B cells [46, 47]. We examined the potential individuals were compared (data not shown). Also, we of HPgV-1 infection to impact on PfSPZ vaccine-induced could not fnd diferences in chemokine and cytokine humoral immunity. PfCSP is the most immuno-domi- levels when comparing the diferent HPgV-1 genotypes nant protein recognized after PfSPZ vaccination. Anti- (data not shown). Taken together, these data suggest that PfCSP titres were measured at baseline in all volunteers (n = 70) and 14 days past last vaccination in all vaccinated Tumbo et al. Virol J (2021) 18:28 Page 9 of 18

Fig. 5 Phylogenetic inferences of the HPgV-1 isolates based on E2 region. Phylogenetic tree was constructed using Neighbour joining method and Kimura two-parameter model of the E2 region of HPgV-1. The E2 sequences from Tanzania and Equatorial Guinea (n 9) were compared to = selected references spanning genotype 1 to 6 from diferent countries available in the NCBI database including; KP701602.1, KM670109, U36380, KP710600, KC618399, AB003291 (Genotype 1, Pink); AF121950, MK686596, D90600 (Genotype 2; 2a Brown), U63715 (Genotype 2; 2b Brown) D87714 (Genotype 3, Green); AB0188667 (Genotype 4, Brown); AY94977, KC618401, AY951979 (Genotype 5, Light blue) and AB003292 (Genotype 6, Green). Equatorial Guinean and Tanzanian strains identifed in this study are denoted by strain number followed with letters EG or TZ, respectively (Red). Chimpanzee HPgV-1 strain (AF70476, Black) was used as outgroup and U4402 (Genotype 2, Golden) was used for mapping of our sequences to identify regions of similarity. The scale bar under the tree indicates nucleotide substitution per site

vaccine was evaluated by presence or absence of asexual volunteers (n = 54) participating in BSPZV2, BSPZV3a and EGSPZV2 (Fig. 7 A-B). Similar results were observed blood-stage parasitemia following homologous PfSPZ when PfSPZ vaccine-induced antibody responses were challenge (CHMI) (Additional fle 3: Fig. 3). While none analysed as net increase (titres at 14 days post last vacci- of the placebo-receiving participants was protected nation minus baseline titres) (Fig. 7 C) and as fold change (0/20), the overall protection in the vaccinated group (titres 14 days past last vaccination divided by baseline was 55% (26/47). Te HPgV-1 prevalence was compara- titres) (Fig. 7 D). Tere was no signifcant correlation ble in these two groups (placebos and vaccines) with 35% between HPgV-1 infection status and anti-PfCSP anti- (7/20) versus 34% (16/47), respectively, suggesting that body titre at baseline and after vaccination. HPgV-1 infection does not facilitate protection against CHMI. We further compared the CHMI protection levels Efect of HPgV‑1 infection on PfSPZ vaccine efcacy in HPgV-1 positive and negative participants in the vac- Te high prevalence of HPgV-1 positive volunteers in cinated group only (Fig. 8a). HPgV-1 positive vaccinees our cohort allowed us to investigate a potential impact showed slightly higher protection levels after CHMI of ongoing viral infection during PfSPZ vaccination on (62.5%; 10/16) compared to HPgV-1 negative individuals vaccine-induced protection. Protective efcacy of the (51.6%; 16/31). We also assessed anti-PfCSP antibodies Tumbo et al. Virol J (2021) 18:28 Page 10 of 18

Fig. 6 HPgV-1 infection is associated with increased systemic levels of IL-2 and IL-17A. Cytokine, chemokine and growth factors levels were analysed by Luminex and levels compared between HPgV-1 negative (5′ UTR-, grey, n 35) and HPgV-1 positive (5′ UTR , purple, n 9) = + = volunteers. Absolute serum concentrations levels (pg/ml) of Interluekin-2 (IL-2) and Interluekin-17A (IL-17A) at samples taken before vaccination are shown. Signifcantly higher IL-2 and IL-17A are seen in the HPgV-1+ compared to the HPgV-1−. Wilcoxon rank sum test was used to determine signifcance (P value * < 0.05) which are indicated on top of top for each group comparison

titres at the peak response which is 14 days past last vac- Efect of CHMI on HPgV‑1 viremia levels cination. Slightly higher anti-PfCSP levels were seen P. falciparum infection is known to impact viremia lev- in protected compared to non-protected individuals els of some common viruses like HIV-1 and EBV [48, (Fig. 8b), without reaching statistical signifcance. PfCSP 49]. We therefore evaluated the efect of an acute P. antibody levels tended to be lower in the HPgV-1 positive falciparum infection on HPgV-1 viremia by comparing individuals (Fig. 8b). the viral load before and 28 days post CHMI in 21 eli- gible trail participants. Detectable levels of HPgV-1 at Interaction of HPgV‑1 and controlled human malaria both time points were quantifed in 9 individuals; 5 of infection induced asexual blood‑stage parasites those showed an increased HPgV-1 viral load and the HPgV-1 co-infection has been associated with favour- other 4 had decreased viremia post CHMI (Fig. 9c). In able clinical outcomes in HIV-1 and Ebola co-infected addition, 12 trial participants were HPgV-1 positive individuals [21–23]. So far, the HPgV-1 impact on P. fal- only for one of the two tested time points; 6 were posi- ciparum infection and immunity is unknown. We evalu- tive before CHMI and 6 were positive at 28 days post ated parasite multiplication rates and pre-patent periods CHMI (Fig. 9c) (Additional fle 4: Fig. 4 showing four in the placebo volunteers only, that have not been vac- representative volunteers individually). cinated (n = 20) undergoing CHMI. Comparable asexual blood-stage multiplication rates and pre-patent periods were observed between HPgV-1 positive and negative individuals (Fig. 9a, b). Tumbo et al. Virol J (2021) 18:28 Page 11 of 18

Fig. 7 HPgV-1 infection status does not impact on anti-PfCSP antibody titres. Total IgG antibodies recognizing full length PfCSP were measured by ELISA in HPgV-1 negative (5′ UTR-, grey), HPgV-1 positive (5′ UTR , purple) volunteers. a Shows baseline (pre-vaccination) anti-PfCSP IgG levels + of HPgV-1 negative (n 47) compared to HPgV-1 positive (n 23) volunteers. b Anti-PfCSP IgG levels at 14 days past last vaccination in HPgV-1 = = positive individuals (n 17) versus the HPgV-1 negative (n 37) group. c, d Comparison of vaccine-induced changes in anti-PfCSP IgG titres as = = net responses (14 days post last immunization—baseline) as well as fold (14 days post last vaccination/baseline). Only vaccinated individuals were included for 14 days post last immunization, net and fold change responses. One HPgV-1+ individual was not included in these subsequent analyses due to missing antibody data. Log anti-PfCSP titres expressed in arbitrary units are shown. Each point represent an individual, box plot with horizontal bar show median values for each group. Statistical signifcance was calculated by using Wilcoxon rank sum test (P value * < 0.05). P values are indicated on top for each group comparison

Discussion was possible only in 2/8 HPgV-1 positive volunteers. Te role of chronic asymptomatic viral infections in Several factors might have contributed to these dis- modulating immune responses in health and disease is crepancies in results, such as the volume and type increasingly appreciated [50]. Te present study sought of samples used. 300 ul of plasma was used as start- to better understand the prevalence and genotype dis- ing material for RNA extraction and RT-PCR, while tribution of HPgV-1 in East and West-central Africa. We roughly 10 times more whole blood was applied as aimed to investigate the potential infuence of HPgV-1 starting material for the RNA-seq data generation. infection on experimental malaria vaccine-induced Also, the variation in HPgV-1 viremia levels between humoral immunity and vaccine-induced protection. serum and cellular components have been demon- By studying a cohort of volunteers undergoing CHMI, strated previously [19, 25]. Te drawbacks of RT-PCR we were in a unique position to investigate if an acute based methods over deep RNA sequencing methods in malaria episode has an impact on HPgV-1 viremia in virus identifcation have been previously reported. To chronically infected volunteers. overcome these limitations, the combination of qPCR In a frst step, we used an unbiased approach by gen- working with predefned primers and unbiased deep erating RNA-seq data to identify prevalent viruses sequencing approaches is recommended [51, 52]. Te circulating in peripheral blood of our BSPZV1 study unbiased virome analyses were important for focusing volunteers that lead to the identifcation of HPgV-1 as our analyses on HPgV-1 in larger cohorts in East and highly present in this group. Confrmation by RT-PCR West Africa. Tumbo et al. Virol J (2021) 18:28 Page 12 of 18

Fig. 8 HPgV-1 infection does not infuence P. falciparum pre-patent periods and parasite multiplication rates during CHMI. Parasitemia was determined in whole blood by qPCR and thick blood smear microscopy (TBS). The analysis included only placebo participants, positive and negative for HPgV-1. a Shows log-fold change of parasitemia in 48 h between HPgV-1 negative (5′ UTR-, grey, n 13) and HPgV-1 positive (5′ = UTR , purple, n 7) volunteers. b Comparison of days post CHMI to malaria positivity by microscopy in HPgV-1 negative (5′ UTR-, grey, n 11) + = = and HPgV-1 positive (5′ UTR , purple, n 7). c HPgV-1 viral loads before (red) and 28 days post CHMI (green) in HPgV-1 infected individuals. Each + = point represents an individual, box plots show data distribution with horizontal bar denoting viral load at each visit. Lines connect viremia levels in individuals found positive for HPgV-1 on both time points. Geometric means were compared between groups and unpaired t-test was used to calculate signifcance. Horizontal bars represent mean with standard deviation Wilcoxon rank sum test was used to compare viremia levels before and after CHMI. P-values are indicated on top of each comparison

Te overall prevalence of HPgV-1 in our main study Similar to a study in Mexico, we observed two broad cohort was 29.2%, roughly the same for Tanzania and groups, low and high, of HPgV-1 viremic individuals, Equatorial Guinea. Te prevalence reported here is likely defned by a cut of value of 600,000 RNA copies/ml. Tis underestimated as we observed fuctuations of HPgV-1 observation likely refects the diferent viral replication viral loads longitudinally, with some volunteers show- states within infected volunteers [22]. We observed simi- ing HPgV-1 positivity in one, two or all three time points lar numbers of high and low viremic individuals, who are assessed. Tese detection variations might indicate either infected with HPgV-1 genotype 2. Te potential role of a viral clearance or continuously ongoing viral replication distinct viral genotype on this pattern remains unclear, with viremia fuctuations sometimes below the lower given the small number of volunteers in this study and detection limit of our RT-qPCR assay [19, 25]. We did not limited heterogeneity of the detected HPgV-1 genotypes. detect HPgV-2 RNA in any of our volunteers but we can- Currently, 7 HPgV-1 genotypes are described globally not completely exclude the possibility of the presence of [54, 55] and some of these genotypes have been impli- circulating HPgV-2 as antibody titers against the HPgV-1 cated in varied clinical outcomes [23, 26, 56]. HIV-1/ and HPgV-2 E2 envelope proteins were not measured. HPgV-1 co-infection studies revealed lower CD4 T cell Our study focussed on the potential impact of HPgV-1 counts in individuals infected with HPgV-1 genotype 2a on PfSPZ vaccine induced humoral responses and pro- than genotype 2b [56, 57] and higher HPgV-1 viral loads tection, thus HIV-1 positive individuals of the BSPZV3a in individuals with genotype 1 compared to genotypes 2a study were excluded from the HPgV-1 association anal- and 2b [58]. Higher serum levels of IFN-ɣ were described yses. It is well known that HIV-1 infection negatively in HIV-1 positive women co-infected with genotype 2 impacts immunity in widely used routine vaccines [53]. compared to genotype 1 [23]. Tumbo et al. Virol J (2021) 18:28 Page 13 of 18

Fig. 9 Association of HPgV-1 infection status with PfSPZ CHMI outcome and anti-CSP titers in immunized volunteers. Individuals were treated with either normal saline (placebo) or PfSPZ Vaccine (vaccinees). Presence or absence of malaria parasites was determined in whole blood by thick blood smear microscopy (TBS) and confrmed by qPCR. Total IgG antibodies recognizing full length PfCSP were measured by ELISA. a Proportion of non-protected (cream) and protected (blue) in vaccinated volunteers with and without HPgV-1 infection. Proportions are indicated inside the bar and volunteer numbers on top. c Total anti-CSP IgG levels at 14 days past last vaccination in the protected (malaria negative) and non-protected (malaria positive) groups, with and without HPgV-1 infection. Log anti-PfCSP titres expressed in O.D units are shown. Each point represent an individual, and box plot with horizontal bar show median values for each group. Chi square with Yates correction was used for group comparisons of categorical values (*P < 0.05). Wilcoxon rank sum test was used to compare anti-CSP titres in the two groups. P values are indicated on top of each comparison

Phylogenetic analyses in our cohort demonstrated than 5′ UTR region like envelope protein 1 (E1), non- the presence of genotype 1 (n = 2, 7.1%) and 2 (n = 24, structural protein 3 (NS3) and non-structural protein 92.3%). Most of our genotype 2 strains clustered within 5A [62, 66]. Te limitations of amplifcation of the 5′ group 2a, originally described from Venezuela. Geno- UTR, a highly conserved region, to discriminate closely type 1 and 2 have been previously reported in Tanzania related isolates is known [67]. Due to its high variabil- but there are no published data available for Equato- ity, E2 provides better genotyping resolution compared rial Guinea [59, 60]. Te predominance of genotype 2 to 5′ UTR. We amplifed and sequenced the E2 region in our study is somewhat surprising. Given the diverse from subjects with high viremia levels in serum (n = 9). geographic origin of our volunteers recruited from East Based on the E2 sequences, these 9 isolates clustered and West-central Africa, we had expected to fnd higher with strains described elsewhere in Africa. It is possible HPgV-1 genetic diversity. Studies in neighbouring coun- that the failure to amplify E2 from all volunteers posi- tries including Cameroon, the Democratic Republic of tive by 5′ UTR detection is either due to the low sen- Congo and Gabon revealed a high prevalence of geno- sitivity of the assay used or the high genetic diversity type 1 [61–65]. Genotypes 2 and 5 were also seen when of the E2 region [42]. While it is known that the detec- phylogenetic studies included molecular markers other tion of HPgV-1 based on amplifcation of the E2 region Tumbo et al. Virol J (2021) 18:28 Page 14 of 18

is highly specifc, it requires higher amounts of RNA context of HPgV-1 infection is unknown. However, in input [42] and individuals with low HPgV-1 viremia are other viral infections like HIV-1 and Hepatitis C, IL-17A likely missed. Alternatively, it is possible that E2 genetic has been shown to promote T-cell mediated anti-viral variants could not be amplifed with the primers used responses through activation and recruitment of den- due to nucleotide sequence mismatch. Te E2 region is dritic cells, monocytes and neutrophils [80, 81]. Other highly variable and this diversity contributes to struc- cytokines and chemokines which could be detected, albeit tural, functional and immunogenic properties of the not signifcantly diferent in HPgV-1 positive individu- virus [68]. Te inconsistent genotyping results of iso- als included SCF (lower) and IL-1beta, IL-12p70, MCP- lates 6EG and 14EG based on 5′ UTR and E2 amplif- 1, LIF, VEGF.A, HGF and TNF-α (higher). BDNF, EGF, cation might be resolved by whole genome sequencing Eotaxin, GRO-alpha, IFN-ɣ, IL-7, IP-10, MIP1-a, Mip-1b, of the virus. Vitrenko et al., reported similar fndings in PDGF.BB, PIGF.1, RANTES, SDF-1a, and VEGF.D were samples from Ukrainian females donating fetal tissues comparable between the two groups. Te levels of these [67]. measured cytokines and chemokines are within ranges Cytokines, chemokines and growth factors are impor- previously reported [26, 27]. While most of the previous tant for inter-cellular communication and regulation of HPgV-1 studies had focused on at risk populations, par- immune processes [69]. Any changes in levels of these ticularly on HIV-1 positive persons, our investigations immune mediators can act as markers of infammation, are in healthy individuals [26, 74], therefor some of the immunity or vaccine uptake [26, 70, 71]. We therefore observed diferences could be due to health status. investigated if altered levels of cytokines and chemokines Here, we observed lower, albeit not statistically signif- unique to ongoing HPgV-1 infection could be identi- cant, median anti-PfCSP titres in the HPgV-1 positive fed. We analysed serum samples taken at baseline for versus the negative group at baseline and 14 days past 45 cytokines in a Luminex platform. Volunteers with last vaccination. Tese observations mirror fndings by chemokine and cytokine levels above the lower limit of Avelino-Silva et al., who showed no association between detection were stratifed according to the HPgV-1 infec- HPgV-1 infection status/viremia with yellow fever spe- tion status. Of all 23 diferentially detected cytokines cifc neutralizing antibody titres in HIV-1 positive indi- and chemokines, IL-2 and IL-17A were signifcantly viduals immunized with yellow fever vaccine [82]. While higher in HPgV-1 positive compared to HPgV-1 negative studies have extensively tried to understand potential individuals. inhibition mechanisms induced by HPgV-1 (and other IL- 2 is an essential survival factor for T and B lympho- Flaviviruses) on T cell activation [77, 83], activation path- cytes [47, 72] and induces the development and survival ways that might be afected in B cells are less explored. of regulatory CD4 T cells critical for the maintenance of It is also possible that the efect of HPgV-1 viruses on immune tolerance [73]. Fama et al., showed increased immune responses against vaccines is negligible when levels of circulating soluble IL-2 receptor (sIL-2R) in studied singly, but this impact is signifcantly synergized HPgV-1 positive volunteers but the authors did not quan- in the presence of other, co-infecting viruses like EBV, tify IL-2 levels [74]. Te increased concentrations of IL-2 CMV and HSV [85, 86]. Hence, the potential role played seen amongst the HPgV-1 positive individuals could by the combined human virome in shaping vaccine- be linked to either on-going antiviral immunity [75] induced responses in diferent populations needs to be or serves as a survival mechanism used by the virus to further explored in larger cohorts. establish persistence in immune cells. A similar mecha- Clinically silent, chronic viral infections are known nism has been described in the apicomplexan pathogen to modulate host immunity [16] and in turn, acute co- Teileria parva that infects T and B lymphocytes in cat- infections are known to drive the re-activation of asymp- tle [76]. Contrary to our observations are results from tomatic viral infections [49]. Several viruses, like HIV-1, HPgV-1/HIV-1 coinfection studies which have shown Ebola and HCV have been implicated in the pathogen- reduced T-cell activation and IL-2 release in coinfected esis and clinical outcome of ongoing malaria infections individuals [77, 78]. Te HPgV-1 envelope protein 2 through a range of diferent mechanisms [84–86]. It has (HPgV1-E2) has been implicated in these outcomes, due been suggested that HIV-1 infections worsen P. falci- to its ability to inhibit T cell-receptor mediated signalling parum presentations by depleting the CD4 T-cell com- and IL-2 signalling pathways [77, 78]. partment, essential for driving malaria-specifc antibody IL-17A induction has been associated with bacterial, responses and for clearance of malaria infected red blood fungal, autoimmune and infammatory diseases [79]. cells [84]. In contrast, better survival outcomes have IL-17A stimulates production of chemokines such as been reported in Ebola infected individuals with P. fal- monocyte chemoattractant protein-1 which mediates tis- ciparum co-infections [85]. Reports have also suggested sue infltration of monocytes. Te role of IL-17A in the delayed emergence of P. falciparum asexual blood-stages Tumbo et al. Virol J (2021) 18:28 Page 15 of 18

in Gabonese individuals chronically infected with HCV Taxonomer and viromescan. C) Viral confrmation: i) Pre-selection criteria [86]. Tus, we studied the impact of HPgV-1 positivity for suspected viral hits by each tool ii) In-silico confrmation of suspected on asexual P. falciparum parasitemia and multiplication viral hits through blasting in NCBI and mapping against specifc viral whole genomes in geneous tool; and removal of viral contaminants. iii) rates during CHMI. Vice versa, we also looked for the Laboratory confrmation of viruses by reverse transcription polymerase frst time at the impact of PfSPZ vaccination and PfSPZ chain reaction. challenge on HPgV-1 viremia. We could not fnd evidence Additional fle 2. Fig. 2 Impact of HPgV-1 infection on systemic cytokines of an association between HPgV-1 infection status and and chemokines. Absolute cytokines, chemokines and growth factor levels at baseline are shown based on HPgV status: HPgV-1 negative (-), asexual blood-stage parasite multiplication rates after grey (n 35) and HPgV-1 positive ( ), purple (n 9). Comparable median CHMI. A slight trend towards longer pre-patent period levels of= Brain derived neutrophil factor+ (BDNF),= Epidermal growth factor was seen in HPgV-1 positive individuals. HPgV-1 positiv- (EGF), Eosinophil chemoattractant cytokine (Eotaxin/ CCL11), Growth regulated oncogene-alpha (GRO-alpha), Interferon gamma (IFN-ɣ), Inter‑ ity appears to increase malaria vaccine-induced protec- luekin-7 (IL-7), Interferon gamma induced protein- 10 (IP-10), Macrophage tion, since slightly higher proportion of CHMI protected infammatory protein 1-alpha(MIP1-a), MIP1-b (Macrophage Infammatory individuals were HPgV-1 positive (62.5% vs 51.6%). How- protein 1-beta), Platelet derived growth factor BB (PDGF.BB), Placental growth factor (PIGF.1), Regulated on activation normal T cells and excreted ever, our study is limited by the sample size and further (RANTES), Stromal derived factor 1 alpha (SDF-1a) , and Vascular endothe‑ investigations with larger cohorts are required to corrob- lial growth factor D (VEGF.D); Lower median levels of Stem cell factor orate these fndings. Importantly, PfSPZ vaccination and (SCF); and higher median levels of Monocyte chemoattractant protein 1 (MCP-1), Leukemia inhibitory factor (LIF), Vascular endothelial growth PfSPZ challenge did not impact HPgV-1 viremia levels in factor A (VEGF.A), Hepatocyte growth factor (HGF) and Tumor Necro‑ our cohort suggesting that the conduct of CHMI is safe sis Factor-alpha (TNF-α) in the HPgV-1 positive individuals. Cytokines, in HPgV-1 infected volunteers. chemokines and growth factors with values above their predefned lower detection limit were considered substantial. Wilcoxon rank sum test was used to compare the two groups and P-values are indicated on top for Conclusions each comparison Notable efects have been reported in HPgV-1 co-infec- Additional fle 3. Fig. 3 Vaccine trial design and procedures. Volunteers are enrolled and randomized into placebo (black icons) and vaccine tions with other RNA viruses such as HIV-1 and Ebola. groups (green icons). Immunized with specifed dose of radiated-attenu‑ Although our study is constrained with limited sample ated whole sporozoites or whole sporozoites with antimalarial drug (V1, size, we have highlighted the epidemiology and genetic V2; V3 etc.) and subsequently challenged with homologous PfSPZ para‑ sites used for vaccination (CHMI). Volunteers are monitored in a controlled distribution of HPgV-1 in areas endemic for malaria. We setting up to 21 days with venous blood drawn daily to monitor presence have reported for the frst time HPgV-1 genotype distri- (malaria positive, not protected) or absence (malaria negative, protected) bution in Equatorial Guinea. We examined the potential of asexual blood-stage parasitemia. All volunteers were treated with an anti-malarial drug either once turning TBS positive or at day 28 after start infuence of HPgV-1 infection status on PfSPZ vaccine- of CHMI. Further monitoring of volunteers occurred at 56 days post CHMI. induced PfCSP-antibody titres and CHMI outcome with- HPgV-1 infection was evaluated in plasma samples from the time points out fnding any striking correlation. Our study provides highlighted in blue. frst time evidence that intravenous vaccination using Additional fle 4. Fig. 4 HPgV-1 RNA positivity and viremia across study visits (Baseline, CHMI and CHMI 28) in Tanzania and Equatorial Guinea. large numbers of attenuated P. falciparum sporozoites HPgV-1 viral plasma RNA was measured+ by RT-qPCR at baseline (pre- and CHMI does not increase HPgV-1 viremia in already vaccination), before (CHMI) and 28 days post immunization (CHMI 28 + infected volunteers. days ) in Tanzanian (n 45) and Equatorial Guinean (n 51) volunteers. Here four volunteers from= the whole cohort are displayed= as a representa‑ tion. The fgure depicts inter-individual variability in HPgV-1 RNA detec‑ Supplementary information tion with some individuals negative or positive at one, two or all three The online version contains supplementary material available at https​://doi. measured time points. Log 10 viral loads are plotted on the y-axis and the org/10.1186/s1298​5-021-01500​-8. time points on the x-axis. Each square plot represents an individual with volunteer identifcation numbers indicated on top. Each dot corresponds to a single time point connected to the next by a solid line. The horizontal Additional fle 1. Fig. 1 Flow chart of volunteers included in virome dashed line indicates the threshold value of zero viremia. pilot study and analyses pipeline. A) Flow chart of volunteers included in virome pilot study and analyses. Samples for transcriptomic studies were Additional fle 5. Table 1 Sensitivity and standard curve ranges for the selected from a subset of volunteers of BSPZV-1 (n 28). RNA sequencing 45 cytokines, chemokines and growth factors analysed in this study. The was performed and, diferential gene expression and= blood transcriptome tables shows the 45 cytokines, chemokines and growth factors their sensi‑ modules were analysed. Non –human reads data was used for virome tivities and standard curve ranges as provided by manufacturer. analyses. B) Virus identifcation: Pilot virome study analysis pipeline- “Bagamoyo viromescan” i) Non-human (un-mapped reads) were searched for “suspected” viral hits in NCBI database containing more than 7424 viral Abbreviations genomes using bowtie 2. ii) Removal of low quality and complexity reads CHMI: Controlled human malaria infection; CSP: Circumsporozoite protein; as well as reads mapping to human genome, transcriptome and repeat E1/2: Envelope glycoproteins (1 and 2); HPgV: Human pegivirus; IVT: In vitro regions by bowtie 2, knead data and tandem repeat fnder algorithms transcription; LEfSe: Linear discriminant analysis efect size; NK: Natural killer respectively. iii) Search for viral hits in the “clean” viral reads using virome cells; NS5A: Non-structural protein 5A; PfCSP: Plasmodium falciparum Circum‑ scan and Taxonomer and for viral proteins using Diamond tool. iv) The sporozoite protein; PfSPZ: Plasmodium falciparum sporozite; PMR: Parasite non-human unmapped reads were also analysed by Fast virome explorer, Multiplication Rate; RNaseP: Ribonuclease P; SSA: Sub-Saharan Africa; TBS: without fltering host reads to allow the identifcation of endogenous Thick blood smear; UTRs: Untranslated regions. retroviral elements and other viruses that may have been missed by Tumbo et al. Virol J (2021) 18:28 Page 16 of 18

Acknowledgements 3. Healer J, Cowman AF, Kaslow DC, Birkett AJ. Vaccines to accelerate The authors would like to thank all participants of the clinical trials for malaria elimination and eventual eradication. Cold Spring Harb Perspect their willingness to participate. We extend our appreciation to the clinical Med. 2017. https​://doi.org/10.1101/cshpe​rspec​t.a0256​27. and laboratory staf involved in these trials (BSPZV1, BSPZV2, BSPZV3a and 4. Langhorne J, Ndungu FM, Sponaas A-M, Marsh K. Immunity to malaria: EGSPZV2), both at the Ifakara Health Institute in Tanzania and the Malaria more questions than answers. Nat Immunol. 2008;9(7):7. https​://doi. Vaccine Initiative in Malabo, Equatorial Guinea. We would also like to thank: org/10.1038/ni.f.205. Prof. Jack Stapleton (University of Iowa), Prof Laurent Kaiser and Dr. Samuel 5. Richie TL, et al. Progress with plasmodium falciparum sporozoite (PfSPZ)- Cordey (University of Geneva) for providing the plasmids which were used based malaria vaccines. Vaccine. Dec. 2015;33(52):7452–61. https​://doi. as positive controls and for generating standards for HPgV strains; Ashery Mbil‑ org/10.1016/j.vacci​ne.2015.09.096. inyi for technical assistance; Tanzania commission of science and technology 6. Seder RA, et al. Protection against malaria by intravenous immu‑ (COSTECH) for providing us with high performance computing facility. nization with a nonreplicating sporozoite vaccine. Science. Sep. 2013;341(6152):1359–65. https​://doi.org/10.1126/scien​ce.12418​00. Authors’ contributions 7. Lyke KE, et al. Attenuated PfSPZ vaccine induces strain-transcending Study concept and design: AT, TS, CD, Investigation: AT, TS, JP, MP, KS, Technical T cells and durable protection against heterologous controlled support and resources: SM, DM, Analyses and interpretation of data: AT, TS, human malaria infection. PNAS. Mar. 2017;114(10):2711–6. https​://doi. NOF, CD, Drafting the manuscript and reviewing: AT, TS, NOF, CD, and all other org/10.1073/pnas.16153​24114​. authors reviewed the manuscript, Study supervision: CD; AO, SJ, Funding 8. Epstein JE, et al. Live attenuated malaria vaccine designed to protect acquisition: CD, MT, KS. All authors read and approved the fnal manuscript. through hepatic ­CD8+ T cell immunity. Science. 2011;334(6055):6055. https​://doi.org/10.1126/scien​ce.12115​48. Funding 9. Jongo SA, et al. Safety, immunogenicity, and protective efcacy against This study was funded by Equatorial Guinea Malaria Vaccine initiative. AT is controlled human malaria infection of plasmodium falciparum sporozoite supported by Swiss government, through ESKAS scheme scholarship Grant vaccine in tanzanian adults. Am J Trop Med Hyg. 2018;99(2):2. https​://doi. No 2016.0056. KS was supported by NIH Grant AI128194. org/10.4269/ajtmh​.17-1014. 10. Jongo SA, et al. Safety and diferential antibody and T-Cell responses to Availability of data and materials the plasmodium falciparum sporozoite malaria vaccine, pfspz vaccine, by Data are available from the corresponding author upon reasonable request. age in tanzanian adults, adolescents, children, and infants. Am J Trop Med Hyg. 2019;100(6):1433–44. https​://doi.org/10.4269/ajtmh​.18-0835. Ethics approval and consent to participate 11. Olotu A, et al. Advancing global health through development and The studies were registered at Clinicaltrial.gov. under the registration numbers clinical trials partnerships: a randomized, placebo-controlled, double- NCT02132299 (BSPZV1), NCT02613520 (BSPZV2), NCT03420053 (BSPZV3a) and blind assessment of safety, tolerability, and immunogenicity of PfSPZ NCT02859350 (EGSPZV2). All clinical trials were approved by the Institutional vaccine for malaria in healthy equatoguinean men. Am J Trop Med Hyg. Review Board for the Ifakara Health Institute (IHI-IRB), Tanzanian Food and 2018;98(1):1. https​://doi.org/10.4269/ajtmh​.17-0449. Drug Administration (TFDA), Tanzanian National Institute for Medical Research 12. Sissoko MS, et al. Safety and efcacy of PfSPZ Vaccine against Plas‑ (NIMR) and the Ethical Committee of Northern and Central Switzerland (EKNZ). modium falciparum via direct venous inoculation in healthy malaria- Written informed consent was obtained from all participants prior enrolment. exposed adults in Mali: a randomised, double-blind phase 1 trial. Lancet All trial procedures were conducted in accordance to good clinical practice Infect Dis. 2017. https​://doi.org/10.1016/S1473​-3099(17)30104​-4. (GCP) and under the Declaration of Helsinki. 13. Hill DL, et al. Immune system development varies according to age, location, and anemia in African children. Sci Transl Med. 2020. https​://doi. Consent for publication org/10.1126/scitr​anslm​ed.aaw95​22. Not applicable. 14. de Bruyn G. Cofactors that may infuence vaccine responses. Curr Opin HIV AIDS. Sep. 2010;5(5):404–8. https​://doi.org/10.1097/COH.0b013​e3283​ Competing interests 3d1fc​a. The authors declare that they have no competing interests. 15. Lenz N, et al. A antiviral innate immune activation in HIV-infected adults negatively afects H1/IC31-induced vaccine-specifc memory CD4 T Author details cells. Clin Vaccine Immunol. 2015. https​://doi.org/10.1128/CVI.00092+​-15. 1 Department of Intervention and Clinical Trials, Ifakara Health Institute, Baga‑ 16. Rocha CS, et al. Subclinical cytomegalovirus infection is associated with moyo, Tanzania. 2 Department of Medical Parasitology and Infection Biology, altered host immunity, gut microbiota, and vaccine responses. J f Virol. Clinical Immunology Unit, Swiss Tropical and Public Health Institute, Socinstr. 2018. https​://doi.org/10.1128/JVI.00167​-18. 57, 4002 Basel, Switzerland. 3 University of Basel, Basel, Switzerland. 4 Equato‑ 17. Rodriguez S, Roussel M, Tarte K, Amé-Thomas P. Impact of chronic viral rial Guinea Malaria Vaccine Initiative, Malabo, Bioko Norte, Equatorial Guinea. infection on T-cell dependent humoral immune response. Front Immu‑ 5 Center for Global Infectious Disease Research, Seattle Children’s Research nol. 2017. https​://doi.org/10.3389/fmmu​.2017.01434​. Institute, 307 Westlake Avenue, N. Suite 500, Seattle, WA 98109, USA. 6 Centre 18. Singh S, Blackard JT. Human pegivirus (HPgV) infection in sub-Saharan Hospitalier Universitaire Vaudois, Lausanne, Switzerland. 7 Iowa City Veterans Africa-A call for a renewed research agenda. Rev Med Virol. 2017. https​:// Administration and the University of Iowa, 200 Hawkins Drive, Iowa City, IA doi.org/10.1002/rmv.1951. 52242, USA. 8 Sanaria Inc, Rockville, MD 20850, USA. 9 Dar-Es-Salaam Institute 19. Chivero ET, Bhattarai N, Rydze RT, Winters MA, Holodniy M, Stapleton of Technology, Dar‑Es‑Salaam, Tanzania. 10 Tanzania Education and Research JT. Human pegivirus RNA is found in multiple blood mononuclear cells Networks, Dar‑Es‑Salaam, Tanzania. 11 Tanzania Commission for Science in vivo and serum-derived viral RNA-containing particles are infectious and Technology, Dar‑Es‑Salaam, Tanzania. in vitro. J Gen Virol. 2014. https​://doi.org/10.1099/vir.0.06301​6-0. 20. Chivero ET, Stapleton JT. Tropism of human pegivirus (formerly known as Received: 24 August 2020 Accepted: 20 January 2021 GB virus C/hepatitis G virus) and host immunomodulation: insights into a highly successful viral infection. J Gen Virol. 2015. https​://doi.org/10.1099/ vir.0.00008​6. 21. Xiang J, et al. Efect of coinfection with GB virus C on survival among patients with HIV infection. N Engl J Med. 2001. https​://doi.org/10.1056/ References NEJMo​a0033​64. 1. Greenwood B. The contribution of vaccination to global health: 22. Horemheb-Rubio G, et al. High HPgV replication is associated with past, present and future. Philos Trans R Soc Lond B Biol Sci. improved surrogate markers of HIV progression. PLOS ONE. 2017. https​:// 2014;369(1645):20130433. https​://doi.org/10.1098/rstb.2013.0433. doi.org/10.1371/journ​al.pone.01844​94. 2. World malaria report 2019. https​://www.who.int/publi​catio​ns-detai​l/ 23. Tillmann HL, et al. Infection with GB virus C and reduced mortality among world​-malar​ia-repor​t-2019. Accessed May 20, 2020 HIV-infected patients. N Engl J Med. 2001. https​://doi.org/10.1056/NEJMo​ a0103​98. Tumbo et al. Virol J (2021) 18:28 Page 17 of 18

24. Lauck M, Bailey AL, Andersen KG, Goldberg TL, Sabeti PC, O’Connor DH. 46. Shibui A, et al. Th17 cell-derived IL-17 is dispensable for B cell antibody GB virus C Coinfections in West African Ebola patients. J Virol. 2015. https​ production. Cytokine. Jul. 2012;59(1):108–14. https​://doi.org/10.1016/j. ://doi.org/10.1128/JVI.02752​-14. cyto.2012.03.018. 25. Rydze RT, Bhattarai N, Stapleton JT. GB virus C infection is associated with 47. Gallou SL, Caron G, Delaloy C, Rossille D, Tarte K, Fest T. IL-2 require‑ a reduced rate of reactivation of latent HIV and protection against acti‑ ment for human plasma cell generation: coupling diferentiation vation-induced T-cell death. Antivir Ther. 2012. https​://doi.org/10.3851/ and proliferation by enhancing MAPK–ERK signaling. J Immunol. Jul. IMP23​09. 2012;189(1):161–73. https​://doi.org/10.4049/jimmu​nol.12003​01. 26. Blackard JT, et al. Cytokine/chemokine expression associated with Human 48. Hofman IF, et al. The efect of Plasmodium falciparum malaria on HIV-1 Pegivirus (HPgV) infection in women with HIV. J Med Virol. 2017. https​:// RNA blood plasma concentration. AIDS. Mar. 1999;13(4):487–94. https​ doi.org/10.1002/jmv.24836​. ://doi.org/10.1097/00002​030-19990​3110-00007​. 27. Lanteri MC, et al. Downregulation of cytokines and chemokines by GB 49. Reynaldi A, et al. Impact of plasmodium falciparum coinfection on Virus C after transmission via blood transfusion in HIV-positive blood longitudinal Epstein-Barr virus kinetics in Kenyan children. J Infect Dis. recipients. J Infect Dis. 2015. https​://doi.org/10.1093/infdi​s/jiu66​0. 2016. https​://doi.org/10.1093/infdi​s/jiv52​5. 28. Nunnari G, et al. Slower progression of HIV-1 infection in persons with GB 50. Gentile G, Micozzi A. Speculations on the clinical signifcance of virus C co-infection correlates with an intact T-helper 1 cytokine profle. asymptomatic viral infections. Clin Microbiol Infect. 2016. https​://doi. Ann Intern Med. 2003. https​://doi.org/10.7326/0003-4819-139-1-20030​ org/10.1016/j.cmi.2016.07.016. 7010-00009​. 51. Bonsall D, et al. Evaluation of viremia frequencies of a novel human 29. Jongo SA, et al. Increase of dose associated with decrease in protection pegivirus by using bioinformatic screening and PCR. Emerg Infect Dis. against controlled human malaria infection by PfSPZ Vaccine in Tanzanian 2016. https​://doi.org/10.3201/eid22​04.15181​2. adults. Clin Infect Dis. Nov. 2019. https​://doi.org/10.1093/cid/ciz11​52. 52. Schlaberg R, et al. Viral pathogen detection by metagenomics and 30. Jongo et. al, “Immunogenicity and protective efcacy of radiation-atten‑ pan-viral group polymerase chain reaction in children with pneumonia uated and chemo-attenuated 4 PfSPZ vaccines in Equatoguinean adults lacking identifable etiology. J Infect Dis. 2017. https​://doi.org/10.1093/ (Jongo et al., manuscript in press).” . infdi​s/jix14​8. 31. Hitchen J, Sooknanan R, Khanna A. Rapid and efcient methods for pre‑ 53. Chaer FE, Sahly HME. Vaccination in the adult patient infected with paring globin- and rRNA-depleted directional RNA-seq libraries. J Biomol HIV: a review of vaccine efcacy and immunogenicity. Am J Med. Apr. Tech. May 2013;24(Suppl):S43–4. 2019;132(4):437–46. https​://doi.org/10.1016/j.amjme​d.2018.12.011. 32. Rampelli S, et al. ViromeScan: a new tool for metagenomic viral com‑ 54. Muerhof AS, Dawson GJ, Desai SM. A previously unrecognized sixth munity profling. BMC Genom. Mar. 2016;17:165. https​://doi.org/10.1186/ genotype of GB virus C revealed by analysis of 5′-untranslated region s1286​4-016-2446-3. sequences. J Med Virol. 2006. https​://doi.org/10.1002/jmv.20510​. 33. Flygare S, et al. Taxonomer: an interactive metagenomics analysis portal 55. Feng Y, et al. A novel genotype of GB virus C: its identifcation and pre‑ for universal pathogen detection and host mRNA expression profling. dominance among injecting drug users in Yunnan, China. PLOS ONE. Genome Biol. 2016. https​://doi.org/10.1186/s1305​9-016-0969-1. 2011. https​://doi.org/10.1371/journ​al.pone.00211​51. 34. Tithi SS, Aylward FO, Jensen RV, Zhang L. FastViromeExplorer: a pipeline 56. Schwarze-Zander C, et al. GB virus C (GBV-C) infection in hepatitis for virus and phage identifcation and abundance profling in metagen‑ C Virus (HCV)/HIV–coinfected patients receiving HCV treatment: omics data. PeerJ. Jan. 2018;6:e4227. https​://doi.org/10.7717/peerj​.4227. importance of the GBV-C genotype. J Infect Dis. 2006. https​://doi. 35. Buchfnk B, Xie C, Huson DH. Fast and sensitive protein alignment using org/10.1086/50571​3. DIAMOND. Nat Methods. Jan. 2015;12(1):59–60. https​://doi.org/10.1038/ 57. Mota LDD, et al. Molecular and clinical profles of human Pegivirus nmeth​.3176. Type 1 infection in individuals living with HIV-1 in the extreme south of 36. G. J. Xu et al., “Viral immunology. Comprehensive serological profling of Brazil. Biomed Res Int. 2019. https​://doi.org/10.1155/2019/80486​70. human populations using a synthetic human virome,” Science, 2015, doi: 58. Giret MTM, et al. Prevalence, incidence density, and genotype distribu‑ https​://doi.org/10.1126/scien​ce.aaa06​98. tion of GB virus C infection in a cohort of recently HIV-1-infected 37. Moustafa A, et al. The blood DNA virome in 8,000 humans. PLoS Pathog. subjects in Sao Paulo, Brazil. PLOS ONE. 2011. https​://doi.org/10.1371/ 2017;13(3):e1006292. https​://doi.org/10.1371/journ​al.ppat.10062​92. journ​al.pone.00184​07. 38. Kearse M, et al. Geneious Basic: an integrated and extendable desktop 59. Stark K, Poggensee G, Höhne M, Bienzle U, Kiwelu I, Schreier E. software platform for the organization and analysis of sequence data. Seroepidemiology of TT virus, GBC-C/HGV, and hepatitis viruses Bioinformatics. Jun. 2012;28(12):1647–9. https​://doi.org/10.1093/bioin​ B, C, and E among women in a rural area of Tanzania. J Med Virol. forma​tics/bts19​9. 2000;62(4):524–30. 39. Frankel M, et al. Development of a high-throughput multiplexed real time 60. Menéndez C, et al. Molecular evidence of mother-to-infant transmission RT-PCR assay for detection of human pegivirus 1 and 2. J Virol Methods. of hepatitis G virus among women without known risk factors for paren‑ Mar. 2017;241:34–40. https​://doi.org/10.1016/j.jviro​met.2016.12.013. teral infections. J Clin Microbiol. Jul. 1999;37(7):2333–6. 40. Krähenbühl S, et al. ELIMU-MDx: a web-based, open-source platform for 61. Smith DB, et al. Discrimination of hepatitis G virus/GBV-C geographical storage, management and analysis of diagnostic qPCR data. Biotech‑ variants by analysis of the 5’ non-coding region. J Gen Virol. 1997. https​:// niques. 2020;68(1):22–7. https​://doi.org/10.2144/btn-2019-0064. doi.org/10.1099/0022-1317-78-7-1533. 41. N’Guessan KF, et al. Human pegivirus (HPgV) infection in Ghanaians co- 62. Tanaka Y, et al. African origin of GB virus C/hepatitis G virus 1. FEBS Lett. infected with human immunodefciency virus (HIV) and hepatitis B virus 1998. https​://doi.org/10.1016/S0014​-5793(98)00083​-0. (HBV). Virus Genes. 2018. https​://doi.org/10.1007/s1126​2-018-1555-2. 63. Liu HF, Muyembe-Tamfum JJ, Dahan K, Desmyter J, Goubau P. High preva‑ 42. Souza IE, et al. Efect of primer selection on estimates of GB virus C lence of GB virus C/hepatitis G virus in Kinshasa, Democratic Republic of (GBV-C) prevalence and response to antiretroviral therapy for optimal Congo: a phylogenetic analysis. J Med Virol. 2000;60(2):159–65. testing for GBV-C viremia. J Clin Microbiol. 2006. https​://doi.org/10.1128/ 64. Tuveri R, et al. Prevalence and genetic variants of hepatitis GB-C/HG and JCM.02663​-05. TT viruses in Gabon, equatorial Africa. Am J Trop Med Hyg. 2000. https​:// 43. Amelio P, et al. HIV infection functionally impairs mycobacterium doi.org/10.4269/ajtmh​.2000.63.192. tuberculosis-specifc CD4 and CD8 T-cell responses. J Virol. 2018. https​:// 65. Luk K-C, et al. Utility of metagenomic next-generation sequencing for doi.org/10.1128/JVI.01728​-18. characterization of HIV and human pegivirus diversity. PLOS ONE. 2015. 44. Douglas AD, et al. Comparison of modeling methods to determine https​://doi.org/10.1371/journ​al.pone.01417​23. liver-to-blood inocula and parasite multiplication rates during controlled 66. Iles JC, et al. Hepatitis C virus infections in the Democratic Republic of human malaria infection. J Infect Dis. Jul. 2013;208(2):340–5. https​://doi. Congo exhibit a cohort efect. Infect Genet Evol. Oct. 2013;19:386–94. org/10.1093/infdi​s/jit15​6. https​://doi.org/10.1016/j.meegi​d.2013.01.021. 45. Segata N, et al. Metagenomic biomarker discovery and explana‑ 67. Vitrenko Y, Kostenko I, Kulebyakina K, Sorochynska K. Prevalence of tion. Genome Biol. Jun. 2011;12(6):R60. https​://doi.org/10.1186/ human pegivirus-1 and sequence variability of its E2 glycoprotein esti‑ gb-2011-12-6-r60. mated from screening donors of fetal stem cell-containing material. Virol J. 2017;14(1):167. https​://doi.org/10.1186/s1298​5-017-0837-y. Tumbo et al. Virol J (2021) 18:28 Page 18 of 18

68. Mohr EL, Stapleton JT. GB virus type C interactions with HIV: the role of 79. Bettelli E, Oukka M, Kuchroo VK. ­TH-17 cells in the circle of immunity and envelope glycoproteins. J Viral Hepat. Nov. 2009;16(11):757–68. https​:// autoimmunity. Nat Immunol. 2007. https​://doi.org/10.1038/ni040​7-345. doi.org/10.1111/j.1365-2893.2009.01194​.x. 80. Meng P, et al. Involvement of the Interleukin-23/Interleukin-17 axis in 69. Borish LC, Steinke JW. 2 Cytokines and chemokines. J Allerg Clin Immunol. chronic hepatitis C virus infection and its treatment responses. Int J Mol 2003. https​://doi.org/10.1067/mai.2003.108. Sci. 2016. https​://doi.org/10.3390/ijms1​70710​70. 70. Domingo-Gonzalez R, Prince O, Cooper A, Khader SA. Cytokines and 81. Yue FY, Merchant A, Kovacs CM, Loutfy M, Persad D, Ostrowski MA. Virus- chemokines in mycobacterium tuberculosis infection. Microbiol Spec‑ specifc interleukin-17-producing CD4 T cells are detectable in early trum. 2016. https​://doi.org/10.1128/micro​biols​pec.TBTB2​-0018-2016. human immunodefciency virus type 1+ infection. J Virol. 2008. https​://doi. 71. Ateba-Ngoa U, et al. Cytokine and chemokine profle of the innate org/10.1128/JVI.02550​-07. and adaptive immune response of schistosoma haematobium and 82. Avelino-Silva VI, et al. CD4/CD8 ratio and KT ratio predict yellow fever vac‑ plasmodium falciparum single and co-infected school-aged children cine immunogenicity in HIV-infected patients. PLOS Negl Trop Dis. 2016. from an endemic area of Lambaréné, Gabon. Malaria J. 2015. https​://doi. https​://doi.org/10.1371/journ​al.pntd.00052​19. org/10.1186/s1293​6-015-0608-4. 83. McLinden JH, et al. Yellow fever virus, but not Zika virus or dengue virus, 72. Ross SH, Cantrell DA. Signaling and function of interleukin-2 in T lympho‑ inhibits T-cell receptor-mediated T-cell function by an RNA-based mecha‑ cytes. Ann Rev Immunol. 2018. https​://doi.org/10.1146/annur​ev-immun​ nism. J Infect Dis. 2017. https​://doi.org/10.1093/infdi​s/jix46​2. ol-04261​7-05335​2. 84. Rénia L, Potter SM. Co-infection of malaria with HIV: an immuno‑ 73. Pol JG, Caudana P, Paillet J, Piaggio E, Kroemer G. Efects of interleukin-2 in logical perspective. Parasit Immunol. 2006. https​://doi.org/10.111 immunostimulation and immunosuppression. J Exp Med. 2020. https​:// 1/j.1365-3024.2006.00903​.x. doi.org/10.1084/jem.20191​247. 85. Rosenke K, et al. Plasmodium parasitemia associated with increased 74. Fama A, et al. Human pegivirus infection and lymphoma risk and progno‑ survival in ebola virus-infected patients. Clin Infect Dis. 2016. https​://doi. sis: a North American study. Br J Haematol. 2018. https​://doi.org/10.1111/ org/10.1093/cid/ciw45​2. bjh.15416.​ 86. Ouwe-Missi-Oukem-Boyer O, et al. Hepatitis C virus infection may lead to 75. Aoshi T, Koyama S, Kobiyama K, Akira S, Ishii KJ. Innate and adaptive slower emergence of P. falciparum in blood. PLoS One. 2011. https​://doi. immune responses to viral infection and vaccination. Curr Opin Virol. org/10.1371/journ​al.pone.00160​34. 2011. https​://doi.org/10.1016/j.covir​o.2011.07.002. 76. C. L. Baldwin et al., “Bovine T cells, B cells, and null cells are transformed by the protozoan parasite Theileria parva.,” Infect Immun, vol. 56, no. 2, Art. Publisher’s Note no. 2, Feb. 1988. Springer Nature remains neutral with regard to jurisdictional claims in pub‑ 77. Stapleton JT, et al. A novel T cell evasion mechanism in persistent RNA lished maps and institutional afliations. virus infection. Trans Am Clin Climatol Assoc. 2014;125:14–26. 78. Bhattarai N, McLinden JH, Xiang J, Kaufman TM, Stapleton JT. GB virus C envelope protein E2 inhibits TCR-induced IL-2 production and alters IL-2–signaling pathways. J Immunol. 2012. https​://doi.org/10.4049/jimmu​ nol.12013​24.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions